De-noising digital radiological images

Abstract

This invention relates to a method for de-noising digital radiographic images based upon a wavelet-domain Hidden Markov Tree (HMT) model. The method uses the Anscombe's transformation to adjust the original image to a Gaussian noise model. The image is then decomposed in different sub-bands of frequency and orientation responses using a dual-tree complex wavelet transform, and the HMT is used to model the marginal distribution of the wavelet coefficients. Two different methods were used to denoise the wavelet coefficients. Finally, the modified wavelet coefficients are transformed back into the original domain to get the de-noised image.

Description

FIELD OF THE INVENTION [0001] This invention relates generally to de-noising digital radiological images. BACKGROUND OF THE INVENTION [0002] General image de-noising techniques based upon the traditional (orthogonal, maximally-decimated) discrete wavelet-transform (DWT) have proved to provide the state-of-the-art in de-noising performance, in terms of peak signal-to-noise ratio (PSNR), according to many papers presented in the literature, e.g. Crouse M, Nowak R, Baraniuk R (1998) Wavelet-based statistical signal processing using hidden Markov models. IEEE Transactions on Signal Processing 46:886-902, Donoho D (1995) De-noising by soft-thresholding; IEEE Transactions on Information Theory 41:613-627, and Romberg J, Choi H, Baraniuk R (2001) Bayesian tree-structured image modeling using wavelet-domain hidden Markov models; and, IEEE Transactions on Image Processing 10:1056-1068. The basic idea behind this image-de-noising approach is to decompose the noisy image by using a wavelet tra...

AI Technical Summary

Benefits of technology

"The invention provides a method for de-noising digital radiological images that improves upon existing methods. The method involves a training stage where a phantom image is transformed using a wavelet transformation to obtain wavelet coefficients, which are then used to estimate parameters of a noise and background distribution. These parameters are then saved for later use in de-noising a subject image. The method can be applied to different body parts and can effectively reduce noise in digital radiological images. The invention also provides a computer-readable memory and an apparatus for carrying out the de-noising method."

Problems solved by technology

Despite the high PSNR values, most of these techniques have their visual performance degraded by the introduction of noticeable artifacts which may limit their use in de-noising of medical images.

The common cause of artifacts in the traditional wavelet-based de-noising techniques is due to the pseudo-Gibbs phenomenon which is caused by the lack of translation invariance of the wavelet method.

Although this problem can be avoided by using an undecimated DWT, it is too computationally expensive.

Images